CN109784279A - X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM - Google Patents
X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM Download PDFInfo
- Publication number
- CN109784279A CN109784279A CN201910045296.1A CN201910045296A CN109784279A CN 109784279 A CN109784279 A CN 109784279A CN 201910045296 A CN201910045296 A CN 201910045296A CN 109784279 A CN109784279 A CN 109784279A
- Authority
- CN
- China
- Prior art keywords
- svm
- ultiwavelet
- power supply
- voltage power
- high voltage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Landscapes
- X-Ray Techniques (AREA)
Abstract
The invention belongs to power failure detection technology fields, disclose a kind of X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM;Multi-resolution decomposition is carried out to collected power work signal using m ultiwavelet tool;Soft-threshold processing is carried out to the m ultiwavelet high fdrequency component coefficient of acquisition, removes the noise in signal;It calculates the sum of absolute value of high frequency coefficient, energy and Energy-Entropy under different scale and normalizes, then various features are combined into new feature vector;Feature vector is sent into trained SVM and obtains final diagnostic result.The present invention overcomes deficiency of the single wavelet in the analysis of power work signal, the SVM algorithm for being combined with stronger generalization ability carries out fault diagnosis, the performance of fault diagnosis is improved, and to realize that the miniaturization of X-ray high voltage power supply, high frequency and industrialization provide technical support and theoretical foundation.
Description
Technical field
The invention belongs to power failure detection technology field more particularly to a kind of X-rays based on multi-wavelet analysis and SVM
High voltage power supply method for diagnosing faults.
Background technique
Currently, the prior art commonly used in the trade is such that X-ray high voltage power supply is the main portion of X-ray electronic system
Part, performance determine that the service life of X-ray application system, significance are self-evident.X-ray high-frequency and high-voltage power supply has
Following advantage: output voltage grade is high, precision is high, temperature drift is small, stability is good, ripple factor is small, small in size, light-weight etc. excellent
Point is therefore widely used in the fields such as medical treatment, science and education, industry, as medical treatment is diagnosed a disease, non-destructive testing, public transport, safety inspection
Etc. various types X-ray machine instrument and equipment in X-ray tube high-voltage power supply special.However, due to high-voltage power circuit collection
Cheng Du, complexity are high, and small fault may cause catastrophic accident, and in addition high voltage power supply high frequency also results in power supply member device
The ghost effect of part is obvious, so that equipment itself is changed into non-linear, time-varying a complication system, it is difficult to establish accurate mathematics
Model is described.Therefore the effective ways for solving high voltage power supply fault diagnosis are explored, On-line Fault control mathematical model is constructed,
Contain failure, ensures equipment safety operation, become the research hotspot and difficult point of current scholars.General X-ray high-voltage electricity
Source fault diagnosis method mainly includes 3 committed steps: being pre-processed firstly for input signal;Then signal is carried out special
Sign is extracted and fusion;Finally to the feature space categorised decision of signal.Currently used Signal Pre-Processing Method has traditional filter
Wave device method or small echo is soft, hard -threshold coefficients model method.There are many ways to feature extraction multiplicity, have by statistical method (based on
Calculate mean value, variance, maximum value and the minimum value etc. of original signal), frequency-domain transform method, deep learning network technique etc..Categorised decision
Means depend in artificial intelligence theory the machine learning algorithm for having supervision, such as: logistic regression, KNN, Adaboost and
SVM etc..SVM supervises two disaggregated models as a kind of classical having, and theoretical based on structural risk minimization, optimal by constructing
Hyperplane obtains globally optimal solution, has stronger generalization ability, not only can solve linearity and non-linearity problem, but also hold very much
Multivariate classification is easily expanded to, therefore using SVM as final fault diagnosis device.Currently based on the analysis target of statistical method
It is original signal, has ignored the time-frequency characteristic of signal, and the excessively single discrimination for leading to signal of feature is too low, it can not be accurate
It is diagnosed to be fault-signal.Fourier transform can disclose the feature of stationary signal as classical time frequency analysis means well,
But the working condition of X-ray high voltage power supply is a unstable, nonlinear complication system, and Fourier transform is not enough to portray letter
Number characteristic, therefore wavelet transformation becomes main analysis means.The small echo being most widely used at present is Daubechies small echo
Series, it can realize multi-resolution decomposition to signal, but single wavelet can not various features waveform in matched signal.
In conclusion problem of the existing technology is:
(1) it is original signal currently based on the analysis target of statistical method, ignores the time-frequency characteristic of signal, and feature
The excessively single discrimination for leading to signal is too low, can not Accurate Diagnosis be out of order signal.
(2) Fourier transform can disclose the feature of stationary signal as classical time frequency analysis means well, but X is penetrated
The working condition of line high voltage power supply is a unstable, nonlinear complication system, and Fourier transform is not enough to portray signal spy
Property, cause diagnostic accuracy lower.
(3) small echo being most widely used at present is Daubechies small echo series, realizes multi-resolution decomposition to signal, point
Frequency extracts signal fault feature, overcome single wavelet can not various features waveform in matching X-ray high voltage power supply working signal, in turn
Improve diagnostic accuracy.
Solve the difficulty of above-mentioned technical problem:
The present invention is the characterization and analysis to X-ray high voltage power supply work wave first, and m ultiwavelet theory is led as mathematics
The more sturdy signal analysis tool of theoretical basis in domain, how the fractional frequency signal in multiple wavelet domains is carried out character representation is this
A maximum emphasis and difficult point for invention.
Solve the meaning of above-mentioned technical problem:
The present invention overcomes deficiency of the single wavelet in the analysis of power work signal, are combined with the SVM of stronger generalization ability
Algorithm carries out fault diagnosis, improves the performance of fault diagnosis, and for realize the miniaturization of X-ray high voltage power supply, high frequency and
Industrialization provides technical support and theoretical foundation.
Summary of the invention
In view of the problems of the existing technology, the X-ray high pressure based on multi-wavelet analysis and SVM that the present invention provides a kind of
Power failure diagnosing method.
The invention is realized in this way a kind of X-ray high voltage power supply fault diagnosis side based on multi-wavelet analysis and SVM
Method, the X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM include:
First by real-time device acquire X-ray high voltage power supply working signal, for signal using GHM m ultiwavelet directly into
Row multi-resolution decomposition obtains each frequency range m ultiwavelet coefficient;
Then it extracts m ultiwavelet high frequency coefficient and carries out soft-threshold contraction, remove noise jamming;
Secondly after the normalization such as each frequency range absolute value of the m ultiwavelet high frequency coefficient after calculation processing and energy and Energy-Entropy
It is spliced into multidimensional characteristic vector;
Characteristic vector is finally sent into trained SVM model, obtains final diagnostic result.
Further, the X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM specifically includes:
Step 1 is the current signal that length is N to input, carries out multi-resolution decomposition to signal using GHM m ultiwavelet, point
The solution number of plies is L, arranges after decomposition according to the number of plies from high to low high frequency coefficient being expressed as DL,DL-1,...,D2,D1, wherein each
Coefficient matrix has 2 dimensions;
Step 2, signal x (n) are indicated are as follows:
X (n)=f (n)+σ z (n), n=0,1 ..., N-1;
F (n) is one-dimensional actual signal, and z (n) is noise, and σ is noise variance, and N is signal length;M ultiwavelet denoises
Reasonable processing is made to m ultiwavelet decomposition coefficient according to practical application;The side of m ultiwavelet soft-threshold processing m ultiwavelet decomposition coefficient
Method expression are as follows:
Wherein ω is wavelet coefficient, and t is threshold value;Threshold value is set
It is high to carry out the m ultiwavelet that L grades obtain after decomposing using GHM m ultiwavelet for step 3, the current signal for being N for length
Frequency coefficient is represented by DL,DL-1,...,D2,D1, the m ultiwavelet coefficient characteristics of extraction include: each frequency band coefficient maximum value,
Minimum value, energy and Energy-Entropy;
Step 4 carries out diagnosis to input feature vector vector using SVM model and specifically includes: SVM and include: Linear SVM and non-
Linear SVM, Linear SVM objective function and constraint condition such as following formula:
The mode of learning of the problem uses Conjugate Search Algorithm, then the dual problem of former problem formula is:
The objective function and categorised decision function of non-linear SVM dual problem are distinguished:
Further, the N in the step 1 and step 3 meets 2 integral number power.
Further, in the step 3:
L layers of high frequency coefficient DLEnergy definition are as follows:
The Energy-Entropy S of m ultiwavelet high frequency coefficient is indicated are as follows:
Further, using SMO algorithm to be decomposed into former quadratic programming problem only in the step 4, there are two the two of variable
Secondary planning subproblem, and Analytical Solution is carried out to subproblem, until all variables meet KKT condition;By didactic
Method obtains the optimal solution of former quadratic programming problem, finally realizes the Efficient Solution of SVM.
Another object of the present invention is to provide the X-ray high-voltage electricity described in a kind of application based on multi-wavelet analysis and SVM
The power failure test platform of source fault diagnosis method.
In conclusion advantages of the present invention and good effect are as follows: the present invention provides one kind to be based on multi-wavelet analysis and SVM
X-ray high voltage power supply method for diagnosing faults, for solve present in existing method for diagnosing faults due to ignore time-frequency spy
Property causes characteristic of division single and single wavelet can not be in matched signal the technical issues of various features waveform.How small use of the present invention is
Wave means carry out Multi-scale Time-Frequency Analysis to working signal, can not only various features waveform in matched signal, and can obtain
Obtain the multidimensional characteristic of signal;It is introduced into SVM algorithm in artificial intelligence theory accurately to classify to the feature vector of input, improves model
Diagnostic accuracy and generalization ability, the present invention can be used for monitoring on-line X-ray high voltage power supply working condition, and then guarantee equipment safety
Operation.
Performance indicator is as shown in table 1 compared with prior art by the present invention, has the advantages that
Multi-scale Time-Frequency Analysis is carried out to signal 1. introducing m ultiwavelet means, it can not only various features wave in matched signal
Shape, and the multidimensional characteristic of signal can be obtained;
2. using the SVM based on structural risk minimization theory as fault diagnosis device, accurately to the feature of input to
Amount classification, improves Model Diagnosis precision and generalization ability.
1 the method for the present invention of table and the prior art carry out performance of fault diagnosis comparison
Detailed description of the invention
Fig. 1 is the X-ray high voltage power supply method for diagnosing faults provided in an embodiment of the present invention based on multi-wavelet analysis and SVM
Flow chart.
Fig. 2 is the X-ray high voltage power supply method for diagnosing faults provided in an embodiment of the present invention based on multi-wavelet analysis and SVM
Implementation flow chart.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to embodiments, to the present invention
It is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not used to
Limit the present invention.
For technical problem of the existing technology, present invention introduces GHM (Geronimo Hardin Massopust) is more
Wavelet analysis means input signal is decomposed and extract each frequency range of high frequency m ultiwavelet coefficient characteristics be sent into it is trained
It is diagnosed in SVM, improves diagnostic accuracy.
Application principle of the invention is explained in detail with reference to the accompanying drawing.
As shown in Figure 1, the X-ray high voltage power supply failure provided in an embodiment of the present invention based on multi-wavelet analysis and SVM is examined
Disconnected method the following steps are included:
S101: multi-resolution decomposition is carried out to collected power work signal (electric current, voltage etc.) using m ultiwavelet tool;
S102: soft-threshold processing is carried out to the m ultiwavelet high fdrequency component coefficient of acquisition, removes the noise in signal;It calculates not
The sum of absolute value with high frequency coefficient under scale, energy and Energy-Entropy simultaneously normalize, then various features are combined into new spy
Levy vector;
S103: feature vector is sent into trained SVM and obtains final diagnostic result.
Application principle of the invention is further described with reference to the accompanying drawing.
As shown in Fig. 2, the X-ray high voltage power supply failure provided in an embodiment of the present invention based on multi-wavelet analysis and SVM is examined
Disconnected method specifically includes the following steps:
Step 1 is the current signal that length is N (N meets 2 integral number power) to input, using GHM m ultiwavelet to letter
Number multi-resolution decomposition is carried out, Decomposition order L arranges from high to low according to the number of plies after decomposition high frequency coefficient being expressed as DL,
DL-1,...,D2,D1, wherein each coefficient matrix has 2 dimensions.
Step 2, it is assumed that signal x (n) can be indicated are as follows:
X (n)=f (n)+σ z (n), n=0,1 ..., N-1 (1)
F (n) is one-dimensional actual signal, and z (n) is noise, and σ is noise variance, and N is signal length.M ultiwavelet denoises
Reasonable processing is made to m ultiwavelet decomposition coefficient according to practical application, then achievees the purpose that denoising.At m ultiwavelet soft-threshold
The method of reason m ultiwavelet decomposition coefficient can be expressed as:
Wherein ω is wavelet coefficient, and t is threshold value.That is the wavelet coefficient less than t sets 0, and the work greater than t becomes 0 processing,
In order to guarantee that the signal after denoising has good adaptability and slickness, threshold value is arranged in the present invention
Step 3 is the current signal of N (N meets 2 integral number power) for length, carries out L fraction using GHM m ultiwavelet
The m ultiwavelet high frequency coefficient obtained after solution is represented by DL,DL-1,...,D2,D1, the m ultiwavelet coefficient characteristics of extraction include: each
Maximum value, minimum value, energy and the Energy-Entropy of frequency band coefficient.
L layers of high frequency coefficient DLEnergy definition are as follows:
The Energy-Entropy S of m ultiwavelet high frequency coefficient can be indicated are as follows:
Step 4, it is described diagnosis is carried out to input feature vector vector using SVM model to specifically include: SVM and include: Linear SVM
And non-linear SVM.Linear SVM objective function and constraint condition such as following formula:
The mode of learning of the problem uses Conjugate Search Algorithm, then the dual problem of former problem formula (5) is:
Needle for the purpose of the present invention, the classification inherently one in X-ray high voltage power supply failure diagnostic process, to feature vector
A Nonlinear separability problem, therefore the present invention mainly studies non-linear SVM.The objective function of non-linear SVM dual problem and point
Class decision function is respectively as shown in formula (7) and (8):
Then, former quadratic programming problem is decomposed by the only quadratic programming subproblem there are two variable using SMO algorithm, and
Analytical Solution is carried out to subproblem, until all variables meet KKT condition.Original two is obtained by didactic method in this way
The optimal solution of secondary planning problem finally realizes the Efficient Solution of SVM.
Application effect of the invention is explained in detail below with reference to emulation experiment.
1. simulated conditions
The present invention is under 10 system of Intel (R) Core (TM) i7-7700CPU@3.6GHz Windows, Matlab
Emulation experiment is completed on 2017a operation platform.
2. experiment content and analysis
For the present invention in the SVM training stage, positive and negative sample size is 150, is tested after training model performance.
100 groups of data are chosen as test sample, first 50 groups are normal waveforms, are artificially introduced corresponding failure in rear 50 groups of data, are utilized
Trained model carries out fault diagnosis.In order to verify effectiveness of the invention, two groups of comparative experimentss and the method for the present invention are designed
It compares, and chooses two kinds of decomposition scales and tested, diagnostic result is as shown in table 2.
2 distinct methods of table carry out performance of fault diagnosis comparison
According to the above experimental result it is found that m ultiwavelet has stronger waveform matching capability, present invention design than single wavelet
M ultiwavelet coefficient characteristics more efficiently characterize different types of waveform characteristic, and the SVM algorithm of use of the invention
It is more effective for X-ray high voltage power supply fault diagnosis.On the whole, the method for the present invention significantly improves the event of X-ray high voltage power supply
Hinder the accuracy rate of diagnosis.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (6)
1. a kind of X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM, which is characterized in that described to be based on
The X-ray high voltage power supply method for diagnosing faults of multi-wavelet analysis and SVM includes:
X-ray high voltage power supply working signal is acquired by real-time device first, signal is directly carried out using GHM m ultiwavelet more
Scale Decomposition obtains each frequency range m ultiwavelet coefficient;
Then it extracts m ultiwavelet high frequency coefficient and carries out soft-threshold contraction, remove noise jamming;
Secondly splice after the normalization such as each frequency range absolute value of the m ultiwavelet high frequency coefficient after calculation processing and energy and Energy-Entropy
At multidimensional characteristic vector;
Characteristic vector is finally sent into trained SVM model, obtains final diagnostic result.
2. the X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM as described in claim 1, feature
It is, the X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM specifically includes:
Step 1 is the current signal that length is N to input, carries out multi-resolution decomposition, decomposition layer to signal using GHM m ultiwavelet
Number is L, arranges after decomposition according to the number of plies from high to low high frequency coefficient being expressed as DL,DL-1,...,D2,D1, wherein each coefficient
Matrix has 2 dimensions;
Step 2, signal x (n) are indicated are as follows:
X (n)=f (n)+σ z (n), n=0,1 ..., N-1;
F (n) is one-dimensional actual signal, and z (n) is noise, and σ is noise variance, and N is signal length;M ultiwavelet denoising is exactly basis
Practical application makes reasonable processing to m ultiwavelet decomposition coefficient;The method table of m ultiwavelet soft-threshold processing m ultiwavelet decomposition coefficient
It reaches are as follows:
Wherein ω is wavelet coefficient, and t is threshold value;Threshold value is set
Step 3, the current signal for being N for length carry out the L grades of m ultiwavelet high frequency systems obtained after decomposing using GHM m ultiwavelet
Number is represented by DL,DL-1,...,D2,D1, the m ultiwavelet coefficient characteristics of extraction include: maximum value, the minimum of each frequency band coefficient
Value, energy and Energy-Entropy;
Step 4 carries out diagnosis to input feature vector vector using SVM model and specifically includes: SVM and include: Linear SVM and non-linear
SVM, Linear SVM objective function and constraint condition such as following formula:
The mode of learning of the problem uses Conjugate Search Algorithm, then the dual problem of former problem formula is:
The objective function and categorised decision function of non-linear SVM dual problem are distinguished:
3. the X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM as claimed in claim 2, feature
It is, the N in the step 1 and step 3 meets 2 integral number power.
4. the X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM as claimed in claim 2, feature
It is, in the step 3:
L layers of high frequency coefficient DLEnergy definition are as follows:
The Energy-Entropy S of m ultiwavelet high frequency coefficient is indicated are as follows:
5. the X-ray high voltage power supply method for diagnosing faults based on multi-wavelet analysis and SVM as claimed in claim 2, feature
It is, using SMO algorithm to be decomposed into former quadratic programming problem only in the step 4, there are two quadratic programming of variable to ask
Topic, and Analytical Solution is carried out to subproblem, until all variables meet KKT condition;Original is obtained by didactic method
The optimal solution of quadratic programming problem finally realizes the Efficient Solution of SVM.
6. a kind of X-ray high voltage power supply failure using described in Claims 1 to 5 any one based on multi-wavelet analysis and SVM
The power failure test platform of diagnostic method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910045296.1A CN109784279B (en) | 2019-01-17 | 2019-01-17 | X-ray high-voltage power supply fault diagnosis method based on multi-wavelet analysis and SVM |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910045296.1A CN109784279B (en) | 2019-01-17 | 2019-01-17 | X-ray high-voltage power supply fault diagnosis method based on multi-wavelet analysis and SVM |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109784279A true CN109784279A (en) | 2019-05-21 |
CN109784279B CN109784279B (en) | 2023-04-18 |
Family
ID=66500966
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910045296.1A Active CN109784279B (en) | 2019-01-17 | 2019-01-17 | X-ray high-voltage power supply fault diagnosis method based on multi-wavelet analysis and SVM |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109784279B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111382790A (en) * | 2020-03-07 | 2020-07-07 | 北京工业大学 | High-voltage direct current fault classification method based on wavelet transformation |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104459388A (en) * | 2014-11-26 | 2015-03-25 | 国家电网公司 | Permanent magnetic direct-drive wind power generation system integrated fault diagnosis method |
CN105956290A (en) * | 2016-05-06 | 2016-09-21 | 国网江苏省电力公司检修分公司 | High-voltage circuit breaker mechanical fault diagnosis method based on multi-data fusion technology |
CN106778592A (en) * | 2016-12-10 | 2017-05-31 | 国网福建省电力有限公司 | Hydropower Unit Condition Monitoring Data noise-reduction method based on m ultiwavelet |
CN107091737A (en) * | 2017-06-06 | 2017-08-25 | 太原理工大学 | A kind of study of typical faults of rotor systems combining diagnostic method based on current signal |
US20180238951A1 (en) * | 2016-09-07 | 2018-08-23 | Jiangnan University | Decision Tree SVM Fault Diagnosis Method of Photovoltaic Diode-Clamped Three-Level Inverter |
-
2019
- 2019-01-17 CN CN201910045296.1A patent/CN109784279B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104459388A (en) * | 2014-11-26 | 2015-03-25 | 国家电网公司 | Permanent magnetic direct-drive wind power generation system integrated fault diagnosis method |
CN105956290A (en) * | 2016-05-06 | 2016-09-21 | 国网江苏省电力公司检修分公司 | High-voltage circuit breaker mechanical fault diagnosis method based on multi-data fusion technology |
US20180238951A1 (en) * | 2016-09-07 | 2018-08-23 | Jiangnan University | Decision Tree SVM Fault Diagnosis Method of Photovoltaic Diode-Clamped Three-Level Inverter |
CN106778592A (en) * | 2016-12-10 | 2017-05-31 | 国网福建省电力有限公司 | Hydropower Unit Condition Monitoring Data noise-reduction method based on m ultiwavelet |
CN107091737A (en) * | 2017-06-06 | 2017-08-25 | 太原理工大学 | A kind of study of typical faults of rotor systems combining diagnostic method based on current signal |
Non-Patent Citations (1)
Title |
---|
李东敏等: "基于多小波包系数熵和人工神经网络的输电线路故障类型识别方法", 《电网技术》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111382790A (en) * | 2020-03-07 | 2020-07-07 | 北京工业大学 | High-voltage direct current fault classification method based on wavelet transformation |
CN111382790B (en) * | 2020-03-07 | 2023-09-29 | 北京工业大学 | High-voltage direct-current fault classification method based on wavelet transformation |
Also Published As
Publication number | Publication date |
---|---|
CN109784279B (en) | 2023-04-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Grezmak et al. | Interpretable convolutional neural network through layer-wise relevance propagation for machine fault diagnosis | |
CN107884706B (en) | Analog circuit fault diagnosis method based on vector value regular kernel function approximation | |
Shi et al. | An independent component analysis classification for complex power quality disturbances with sparse auto encoder features | |
CN112036301B (en) | Driving motor fault diagnosis model construction method based on intra-class feature transfer learning and multi-source information fusion | |
CN109974782B (en) | Equipment fault early warning method and system based on big data sensitive characteristic optimization selection | |
CN102122133B (en) | Self-adaption wavelet neural network abnormity detection and fault diagnosis classification system and method | |
CN101587155B (en) | Oil soaked transformer fault diagnosis method | |
Hajian et al. | A new hybrid pattern recognition scheme for automatic discrimination of power quality disturbances | |
CN108573225A (en) | A kind of local discharge signal mode identification method and system | |
CN102435910B (en) | Power electronic circuit health monitoring method based on support vector classification | |
CN110705456A (en) | Micro motor abnormity detection method based on transfer learning | |
CN110348491A (en) | Rolling bearing fault recognition methods based on study dictionary and singular value decomposition | |
CN110298085A (en) | Analog-circuit fault diagnosis method based on XGBoost and random forests algorithm | |
CN110456217B (en) | MMC fault positioning method based on WPD-FOA-LSSVM dual model | |
CN109323754A (en) | A kind of train wheel polygon fault diagnosis detection method | |
CN110070102B (en) | Method for establishing sequence-to-sequence model for identifying power quality disturbance type | |
Lima et al. | Wavelet‐artificial immune system algorithm applied to voltage disturbance diagnosis in electrical distribution systems | |
CN104951787A (en) | Power quality disturbance identification method for distinguishing dictionary learning under SRC framework | |
CN110533007A (en) | A kind of vehicle-mounted strain of bridge influences the identification of line feature intelligent and extracting method | |
CN202119467U (en) | Self-adaptive wavelet neural network categorizing system of anomaly detection and fault diagnosis | |
CN113869145A (en) | Circuit fault diagnosis method and system for light-weight gradient elevator and sparrow search | |
CN114700587B (en) | Missing welding defect real-time detection method and system based on fuzzy inference and edge calculation | |
CN114237046A (en) | Partial discharge pattern recognition method based on SIFT data feature extraction algorithm and BP neural network model | |
CN114081508B (en) | Spike detection method based on fusion of deep neural network and CCA (common cancer cell) characteristics | |
Guzmán et al. | Wavelet denoising of partial discharge signals and their pattern classification using artificial neural networks and support vector machines |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |